Regularization Reconstruction Method for Imaging Problems in Electrical Capacitance Tomography

2017 3rd International Conference on Energy, Environment and Materials Science

能源学;生态环境科学;材料科学

Chu, Pan^1,2 ; Lei, Jing^3

Guangdong Electric Power Design Institute Co., Ltd., China Energy Engineering Group, Guangzhou

510663, China^1

Institute of Electrical Engineering, Chinese Academy of Sciences, Haidian District, Beijing

100190, China^2

School of Energy, Power and Mechanical Engineering, North China Electric Power University, Changping District, Beijing

102206, China^3

关键词: Efficient numerical method;    Electrical Capacitance Tomography;    ILL-posed inverse problem;    Multi-phase flow systems;    Parametric measurements;    Regularization reconstruction;    Tikhonov regularization;    Visualization measurements;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/94/1/012127/pdf
DOI  :  10.1088/1755-1315/94/1/012127

来源: IOP

【 摘 要 】

The electrical capacitance tomography (ECT) is deemed to be a powerful visualization measurement technique for the parametric measurement in a multiphase flow system. The inversion task in the ECT technology is an ill-posed inverse problem, and seeking for an efficient numerical method to improve the precision of the reconstruction images is important for practical measurements. By the introduction of the Tikhonov regularization (TR) methodology, in this paper a loss function that emphasizes the robustness of the estimation and the low rank property of the imaging targets is put forward to convert the solution of the inverse problem in the ECT reconstruction task into a minimization problem. Inspired by the split Bregman (SB) algorithm, an iteration scheme is developed for solving the proposed loss function. Numerical experiment results validate that the proposed inversion method not only reconstructs the fine structures of the imaging targets, but also improves the robustness.